Interactive comment on "Decennial time trends and diurnal patterns of particle number concentrations in a Central European city between 2008 and ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Atmos. Chem. Phys. Discuss.,
https://doi.org/10.5194/acp-2020-305-RC2, 2020 ACPD
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
Interactive
comment
Interactive comment on “Decennial time trends
and diurnal patterns of particle number
concentrations in a Central European city between
2008 and 2018” by Santtu Mikkonen et al.
Anonymous Referee #2
Received and published: 7 August 2020
General comments
The paper presents convincing arguments with statistical methods as why it is likely
that decreased traffic emissions have led to decreased particle pollution in Budapeszt.
Hence, these results are extremely important from a policy making point of view. In
other words, it is neccessary that there are no doubts as to why concentrations have
gone down. For this reason, the paper must undergo additional analysis to prove the Printer-friendly version
point that car emissions have indeed led to the decrease of particle concentrations in
Budapeszt. If this analysis can not conclusively prove the reason for the decrease, Discussion paper
then I must stress the need to be more careful about the conclusion in the abstract
C1and conclusion section as to the reason for the decrease. And provide a sentence that
other reasons for the decrease can not be ruled out. ACPD
Hence, I suggest to accept the paper subject to major revision of data analysis as
suggested below. In addition there are a few minor remarks that need to be addressed.
Interactive
Major remarks comment
Table 2 seems to indicate that there are increasing number of cars and car age as
function of time in Budapeszt. How this can lead to decrease in emissions from traffic
in chapter 3.1 is not quite convincing despite that Euro regulations are imposing re-
strictions to emissions from new cars. The total emissions are likely dominated by the
ageing cars, and the average age of cars is increasing with time. Is there still a like-
lihood that the statistical analysis could be wrong for some unknown reason, and the
reason for decreasing concentrations is arising as function of varying meteorological
conditions between different years as expressed by the MCP parameter and not due
to decreased car emissions? After all, the gas concentrations are not decreasing with
time as the authors admit. I want the authors to quantify how often MCP weather pat-
terns of type 3, 7, and 12 occur during April, May and August during the earlier years
compared to the later years, and see if this can explain why the N6-1000 is higher in
the earlier years than the later years (and also for the MCP that do not favor NPF, type
6).
The statistical method of DLM and GLMM are not enough for interpreting the results.
A manual analysis of the MCP as described above must be done as well to provide
further evidence for decreasing particle trend as being caused by decreasing traffic
emissions. I stress also to use an educated guess/calculation of how you expect to
see reductions in emissions of particles based on the values of car age and number Printer-friendly version
of vehicles and diesel car share as function of year, and see if your expecations in-
deed would indicate a decrease in emissions. Just because there is a tendency for Discussion paper
decreasing car emissions in Germany, doesn’t prove that the same thing is happening
C2for Budapeszt if the traffic fleet composition is different in Budapeszt. Also, I want the
authors to look more carefully on meteorological parameters manually as well. Even ACPD
though for example the average temperature is the same for the different years, doesnt’
mean that it is not varying between different days that lead to the particle concentration
trends between different years. Interactive
comment
Minor remarks
Please denote that the station is an ”urban background” station in the abstract, intro-
duction and methodology sections. This is needed for other readers to relate to the
expected pollution level, and to know if this is the most polluted place in Budapest, or
as you have in this case, a medium population exposure location in the city center, so
called urban background. Alternatively, if it is not a typical urban background site, but
slightly more polluted (I don’t know this), please explain in the introduction and method-
ology section that the site has a pollution level between a typical heavy trafficked street
level site and an urban background site, but closer to typical urban background levels.
Lines 133-134. ”the most extensive inter-comparison was realised in summer 2015
(Salma et al., 2016a) and autumn 2019”. Please denote which kind of intercomparisons
were made.
Chapter 2.1: SO2, CO, NO, NOx, O3, and PM10 measurements: What is the pollution
level of the site measuring these paramters: urban background or street level pollution,
or something inbetween, or cleaner than an urban background site? Please describe
this station as well. Otherwise, we cannot compare the time trends for this site as
compared to the BpArt site. For example, the BpArt site might be closer to traffic than
the gas measurement site explaining why the BpArt concentrations are decreasing with
decreasing traffic emission trends, but not at the gaseous concentration site, which is Printer-friendly version
then relatively more influenced by background long range sources.
Discussion paper
Chapter 2.2: The MCP codes are developed for 00:00 UTC time. When you have a
time of your particle or gas concentration data or meteorological data, which is the
C3MPC type that you use? For example, if the measurement time is August 1, 14:00 local
time, what is the MCP coding for that time? Is it the MCP coding from August 1, 00:00 ACPD
UTC, or MCP coding from August 2, 00:00? Or is it denoted as a combination of both
MCP codes? It should be clearly stated in the manuscript. The MCP coding from one
day to the other might change completely, meaning MCP codes for a measurement Interactive
time inbetween two MCP coding times can be ambiguous. comment
Chapter 2.3.1. Would you please explain the autoregressive component?
Equation (4): The MCP is not a continuous variable, but it is discrete. How can you
construct a linear output factor from Beta-6 multiplied with MCP-i? Would you mind
explaining how Beta-6 and MCP-i and their product are constructed?
Line 274: Q (GRad) calculation is incorrect. The equation is correct only if you have
100 % data coverage. You have between 90 and 100 % data coverage as indicated in
the method section. Hence, the calculated value will be systematically underestimated
unless you interpolate data for the missing hours of GRad data. This could potentially
be the reason why Table 2 Q values are different for different years, and not due to
varying total insolation during one year to the other.
Lines 317-318. It is a strong statement to say that ”this decoupling confirms that the
causes of the decrease in particle number concentrations are not primarily related to
meteorological conditions because they would jointly affect the gas concentrations as
well”. That gas concentrations don’t go down and particle number concentrations go
down could by accident also be related to a difference in availability of different MCP
days during different years and seasons. NPF events could be favored in earlier years
due to for example quite high number of certain MCP days with lower particle surface
area, which favours NPF, which don’t appear as frequent in the later years. This could Printer-friendly version
happen even if the median particle surface area is decreasing every year (as indicated
by decreasing N100-N1000). But, this difference in MCP does not automatically mean Discussion paper
that the gas concentrations should change in the same way as N6-1000. Hence, I
C4would rephrase the wording from ”confirms” to ”suggests”.
Conclusion: You mention that the accumulation mode particles don’t show a decreas-
ACPD
ing annual trend. But, according to Table 3 they do.
Interactive
Interactive comment on Atmos. Chem. Phys. Discuss., https://doi.org/10.5194/acp-2020-305,
comment
2020.
Printer-friendly version
Discussion paper
C5You can also read
